The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external transaction descriptions, which theoretically could contain malicious instructions.
Ingestion points: Data enters via the transaction_search tool results as described in the Workflow.
Boundary markers: The instructions do not specify the use of delimiters or warnings to ignore instructions embedded in the transaction data.
Capability inventory: The skill is limited to performing transaction searches and formatting user-provided data into a text statement.
Sanitization: There is no mention of sanitizing or filtering the content of transaction descriptions before use.

net-worth